Leo 🇦🇱 🇦🇷(UE Cornellá fan 💚 and CD Eldense)

Jensen Huang nearly lost NVIDIA three times before anyone cared. First chip (NV1) bet on the wrong graphics math. Microsoft standardized on triangles. Almost everything shipped came back. Cash was measured in weeks. He flew to Sega, admitted their next chip was a dead end, and asked them to cancel the contract but still pay the remaining money. They did. That cash bought one more shot. RIVA 128 was designed almost entirely in simulation. No money for respins. It worked. A million units in four months. Company survived. Years later, Intel used a licensing dispute to push NVIDIA out of the chipset business entirely. Jensen took the hit, exited, and doubled down elsewhere. He then spent a decade building CUDA so GPUs could do more than graphics, for a market that barely existed. Wall Street hated it. He kept going. When AI arrived, the architecture and software he had forced through were already there. One man absorbed the near-death cycles, the public failures, and the long years of no payoff, then delivered the platform the world is now running on. That level of resilience, pain tolerance, and vision is rare. He should keep running NVIDIA.